Co-optimization of codon usage and mRNA secondary structure using quantum computing

TL;DR

This paper presents a novel quantum computing-based framework for simultaneously optimizing mRNA sequences for codon usage and secondary structure, improving stability and efficacy of mRNA therapeutics.

Contribution

It introduces a dual-objective variational approach that combines codon optimization and secondary structure prediction using hybrid quantum-classical algorithms.

Findings

Feasibility demonstrated on IBM's 127-qubit quantum hardware.

Validated approach with simulations and hardware experiments on sequences up to 30 nucleotides.

Potential to accelerate mRNA therapeutic design processes.

Abstract

Co-optimizing mRNA sequences for both codon optimality and secondary structure is crucial for producing stable and efficacious mRNA therapeutics. Codon optimization, which adjusts nucleotide sequences to enhance translational efficiency, inherently influences mRNA secondary structure - a key determinant of molecular stability both in-vial and in-cell. Because both properties are governed by the same underlying sequence, optimizing one directly impacts the other. To address this interdependence, we introduce a novel variational framework that simultaneously optimizes codon usage and secondary structure. Our method employs a dual-objective function that balances the codon adaptation index (CAI) and minimum free energy (MFE), incorporating variational parameters for codon selection. Leveraging a hybrid quantum-classical computational strategy and building on prior advancements in quantum algorithms for secondary structure prediction, we effectively navigate this complex optimization space. We demonstrate the feasibility of executing this end-to-end workflow on real quantum hardware, using IBM's 127-qubit Eagle processor. We validate our approach through both simulations and quantum hardware experiments on sequences of up to 30 nucleotides. These results highlight the potential of our framework to accelerate the design of optimal mRNA constructs for therapeutic and research applications.